314 ALASKAN GLACIER STUDIES and the glaciers. In outer College Fiord near Port Wells and on the flats near Amherst Glacier, there is thick, mature spruce and hemlock forest. On the eastern side the forest extends northward nearly to Yale Glacier, ending at sea level in the cove three miles southwest of the glacier. From this point northward there is no forest at sea level, though scattered thickets of .willow and alder have reached considerable size. On the eastern mountain side, however, mature spruce extends nearly up to the first tributary of Yale Glacier. It is limited above by a nearly-horizontal timberline, but its lower edge is a rather regular, sloping boundary1 1017 feet above sea level near the glacier and 200 feet in the cove to the southwest, practically coinciding with deposits thought to represent a faint, disconnected lateral moraine. The western side of Yale Arm has a similar distribution of mature, thickset timber, whose growth must have taken a century or more. At the end of College Point there are scattered individuals of mature spruce at sea level, while the thick, mature forest begins at an elevation of about 575 feet, rising northeastward, with the lateral moraine already described, to 900 feet near the western edge of Yale Glacier. Above this line the forest extends up the mountain, slope to the timberline; but below it are only scattered young conifers and dense thickets of aider and willow up to and beyond the end of the glacier, In these lower levels there is no thick mature evergreen forest such as extends up to the very edges of Columbia and Meares Glaciers. Some of the alders close to the front of Yale Glacier are S3 years old and the oldest willow seen had ten annual rings. From this statement it is evident that the forest conditions in Yale Arm of College Fiord are unusual, for there is a wedge of thick, mature, coniferous forest on each side, extending upward to a normal but somewhat irregular timberline, and limited below by a regular line, highest near the glacier and sloping gradually southwestward to sea level. The questions naturally arise: Is the absence of forest in the lower part of this area to be explained by unsuitability of (a) the slopes, or (b) the soil? Or is it due to a difference in the length of time during which this area has been available for tree growth? The last seems to be the case, for in the southern part of College Fiord there is thick, mature forest on all sorts of slopes and soils. The cause which until recently has kept the forest from growing in this lower belt was clearly the Yale Glacier, for an advance to or near the mouth of Yale Arm would cover the fiord walls up to about the present lower limit of the forest. The unfcrested area is, therefore, interpreted as a former barren zone, already thickly covered with alders and willows, some of which are S3 or more years old, while scattered young conifers have taken root where slope, soil, and adjacent seeding ground are most favorable. This interpretation is based upon our belief that the upper belt of thick, mature forest is a remnant of a forest which formerly clothed the entire lower portion of the fiord wall, and not a forest that grew where it is while the glacier occupied the area below it. Such remnant forests exist to the west in Harriman Fiord, and we are inclined to interpret thia one as of the same character, in spite of the fact that in this. case we saw no push moraines with dead tree trunks at the lower and outer edges of the forest. The vegetation in Yale Arm, therefore, is interpreted as showing that the glacier extended out nearly to the end of College Point not less than 88 years ago; and, furthermore, that for a century or more it has not extended farther than that, nor higher than 900 to 1017 feet above the present terminus. 1 Not Horizontal, as Grant thought. Op. cit., p. 324.